Ambient temperature, sunlight duration, and suicide: A systematic review and meta-analysis

The relationship between sunlight hours, temperature, and suicide has been investigated in many previous studies, but with inconsistent findings. We conducted this meta-analysis aiming at providing clear evidence about whether sunlight hours and temperature can affect the risk of suicide.

Relevant literature published before 7 Jul 2018 was searched in three electronic database of PubMed, Web of science and Cochrane. Between-study heterogeneity in the effect estimates were examined by study design (i.e., exposure time resolution, temperature measure and suicide classification) and population vulnerability (i.e., sex, national income level, climate zone and study location).

Regarding temperature-suicide association, we identified 14 studies that provided 23 effect estimates for meta-analysis. Our random-effects model showed that each 1 °C increase in temperature was significantly associated with a 1% increase in the incidence of suicide (Incidence Rate Radio (IRR) = 1.01; 95%CI = 1.00-1.02; p < 0.05). As for sunlight duration and suicide, 11 effect estimates from 4 studies were pooled using a fixed-effects model. Significant association between sunlight duration (per 1 h increase) and the risk of suicide was not observed, and the pooled IRR was 1.00 (95%CI = 0.99-1.01; p > 0.05). There was significant heterogeneity between previous studies' effect estimates, and significantly lower heterogeneity was noted for the female group, for studies that did analyses at daily scale, and for studies using mean temperature as the exposure indicator.

Our findings suggest a significant and positive association between temperature rises and incidence of suicide, and a non-significant association between sunlight duration and incidence of suicide. Populations living in tropical and temperate zones or at middle-income level may be particularly vulnerable to temperature increase.